We need to talk about the National Academy of Sciences

Last week, the President of the National Academy of Sciences (NAS) came out with an opinion article in PNAS (operated by the NAS). The opinion is “Plan S falls short for society publishers—and for the researchers they serve.” 

Missing from the opinion was an outline of what NAS has to gain if “Plan S” is not fulfilled. And while it went into some detail on how, in general, societies would lose out with “Plan S”  (with arguably questionable math), it didn’t explain why the NAS in particular fears it would have losses. A look at its financial statements shows an organization that is grossly overspending on non-charitable activities, even when benchmarked against similar non-profits. Out of every $1 the NAS spends, only $0.17 is going back out to grants and similar (non-research/mission related overheads are the bulk of remaining expenses - see figure 1). While the Gates Foundation gives back $0.84 of every dollar spent, and Wellcome Trust (UK) returns $0.71 of every dollar equivalent spent.

Of these three organizations, only the latter two (Gates and Wellcome) support PlanS and Open Access. One really has to wonder why the huge discrepancy in charitable spending, and whether there is a connection to the support for or against Open Access and PlanS. Taken to the extreme, if PlanS forced the NAS to close down (unlikely event, of course), then more money would in theory be available for research grants, as more efficient charities could take charge.

It feels like blasphemy to be saying that research could possibly be more well off if the National Academy of Sciences ceased to operate. It feels like heresy, because unlike most non-profits, the NAS was formed under a congressional charter. Other than its beginnings, the NAS is no different than other non-profits, however. It is independent of congress, just like all other corporations and non-profits. A congressional charter confers no special privilege or status, it is merely symbolic. Congressional charters for new non-profits were in fact stopped several decades ago. The NAS, like all other non-profits, has no official governmental function in the US, although the US government can and does call up the NAS for scientific direction (the US does this with other for-profits and non-profits as well).

What the congressional charter’s designation, history and naming has uniquely conferred upon the NAS is a legacy halo-effect with little oversight into how well it is operating internally. Like many bloated non-profits, it has painted itself into a corner where the cost of overheads and legacy spending habits has removed any ability to look to the future. It has become the proverbial deer in headlights too afraid to progress; Open Access and PlanS are the cars driving into the future about to run it over. 

If the NAS fears it cannot survive this future, then what in the hell is it good for and what should be done about it? The problem seems to be the poorly ran NAS, rather than Open Access. The NAS looks like it is barely capable of surviving any environmental changes that it faces going forward. This is before we get into the irony of an organization with the mission of “furthering science in America,” and yet uses its own publications and platform that others will look to (without questioning) to shout down Open Access and other “challenges” it fears.

When the NAS says it has concerns about Open Access and PlanS, what it really means is that its OWN tenuous financial stability is being threatened. This is not the same as threatening all of science, as we see with Wellcome, Gates, and other organizations that are able to thrive. The NAS mission also states that it is supposed to provide “objective advice to the nation on matters related to science.” If the NAS has become financially compromised, can we really count on it to remain “objective?” 


Figure 1 National Academy of Sciences financials (Form 990 2016 most recent). Previous years are similar in terms of proportion of expenditures towards various activities.

The NAS had a total expenditure of $312.4M (line 19) in 2016 from the most recent available financials. Of this amount, $53.8M was spent on grants and other charitable activities (line 13). This amounts to $0.17 of each dollar spent going towards grants, etc (or $0.14 of each revenue dollar). Meanwhile $132.4M was spent directly on salaries and benefits (or $0.42 of each dollar spent). Line 17 shows “other expenses” totaling $126M. That sum includes $19.95M in travel expenses, $18.3M in “occupancy” or building rents, $13.2M on IT and so on. Only $8.4M went towards “Conferences, conventions, and meetings.”


Figure 2 Wellcome Trust financials for 2018

Digging into Wellcome’s expenditures the available financials are more recent, 2018. Wellcome spent £781.2M in 2018, of which £638.1M went toward grants and research, etc. Of that £638M, £86.2M was allocated for support, which includes all overheads (salaries, travel, rents, etc). That left £551.9M for direct charitable activities, or $0.71 per dollar spent.


Figure 3 Gates Foundation financials for 2016 (most recent year available)

Above, Figure 3 line 12 shows total revenues for the Gates Foundation of $5.28B (that’s billion) and expenses (line 26) of $5.7B. Of the expenses, $4.847B went toward grants, etc (or $0.84 of every dollar spent).


Figure 4 Wellcome Trust employee compensation

I didn’t want to get into individual salaries or compensation in the main paragraphs at the top. First, these are huge organizations and while salaries can be a lot, it isn’t clear that highly paid executives directly contributes towards the massive inefficiencies seen in the NAS. The 15 highest paid executives earned $7.8M in compensation, including $1.1M to the President of NAS. Of total expenditures, that is 2.5% for the highest paid executives ($132.4M for all staff or 42%). The Gates Foundation has similar levels of executive compensation, BUT as a total proportion of expenses its total staff expenditure is only 5.7% of all costs. The inefficiencies at the NAS are seen only when looking at all staff costs. 

Meanwhile, figure 4 shows staff compensation (salary, bonus, etc) to Wellcome Trusts employees (it excludes the investment team, which has performance based salaries depending on investment income). Again, executive compensation isn’t the real issue, although the highest paid Wellcome directors are paid just half of their US counterparts at NAS. The real story is the efficiency of total staff expenditures. 


The (usually) forgotten advice in startup advice

I’ve read two seemingly diametrically opposed startup retrospectives this week. Really, same goes for every week.

The first was a post-mortem of a startup failure that tried making customized and crowd-sourced designer jeans.


The second was a lesson on getting turned down by VCs from Justin Kan, founder of Twitch, which lets people watch others playing video games.


In the first instance, the moral of the story as told by the founder is to do your homework, i.e. market research and validation before starting the business full-on and wasting six years of your life.

In the second instance, Twitch, the moral of the story is to not listen to anyone who thinks your idea is crazy and persevere through it. Show the world!

Here’s what these two blog posts (and similar) usually forget to mention…

Both of those bits of advice are wrong and correct. Sometimes you can’t validate an idea ahead of time. Other times you can validate market potential, but the data may tell you to quit, when actually it could very well succeed. Or it could be the opposite, the market says yes this is a winner, and it turns out not to be. 

So what the hell is one to do then?!

Really, there’s only one thing you can do - if your idea is uncertain and you decide to follow through with it, then become comfortable with that uncertainty. If you don’t have that risk appetite, then head the other direction and look for an idea with more certainty in the market (which will have different barriers of course).   

Not every good idea is worth advancing, and not every crazy idea should be dropped. The forgotten piece of advice is that you need to ask yourself how much risk you’re willing to take on board. And that’s unique to everyone. 

Even for the seasoned and experienced, that can be a difficult question to honestly answer. So, look for parallels in your life where you were possibly biting off more than you could chew, its outcome, and how you felt about it. Validating your appetite for risk is probably the most important element to starting something. There’s risk in any business venture.


Slack API users allowing gov’t surveillance are in breach of Slack terms!

Just spotted this as Slack has recently updated their API Terms of Service, though it may have been in there before this update?

Of course, faced with either jail time courtesy of the government or breaking Slack’s TOS, which do you think a developer would choose? Slack legal time just covering their arse with this clause.


9.1 Government Access: You will not knowingly allow or assist any government entities, law enforcement, or other organizations to conduct surveillance or obtain data using your access to the Slack API in order to avoid serving legal process directly on Slack. Any such use by you for law enforcement purposes is a breach of this API TOS.


Download academic PDFs? You might get punched in the face

Courtesy of JSTOR. Well, hopefully not JSTOR as an organization, but at least one of their employees thinks you should get knocked out. That was just learned from a Freedom of Information Act request obtained from the Aaron Swartz saga. You can view all of the files obtained from the FBI, Secret Service, MIT, JSTOR, US Attorney’s Office and others at swartzfiles.com


The above is from page 4 of more than 3K pages of internal emails from ITHAKA.org, which is the non-profit responsible for the JSTOR digital library (whose mission is “to foster widespread access to the world’s body of scholarly knowledge”). It’s a couple of systems administrators communicating in the context of high download activity going on at MIT (which we later learn was Aaron Swartz downloading academic papers).

Sadly, despite its mission, JSTOR  believes it is no longer capable of sustaining itself in the digital era without resorting to restricting access to knowledge. The Internet and World Wide Web, designed to spread information, have changed everything, sometimes ironically.

For its part, JSTOR settled a civil suit with Aaron Swartz out of court, and later told the US Attorney’s office in Massachusetts that it no longer had an interest in further proceedings. The US Gov’t didn’t stop, however.

Getting punched in the face was the least of Aaron’s worries. The FBI had much worse in mind for him (and succeeded in doing) for the act of downloading PDFs. It culminated in the unnecessary loss of his life.

There’s too much in the FOIA for any one person to really go through entirely, but have a perusal of the documents and let’s remember what Aaron stood for, and that we can do better when it comes to being stewards of the world’s academic knowledge. As I’ve stated before on this blog, without Aaron there may have never been the motivation to start the journal PeerJ. News of Aaron in 2011 was the catalyst to finally say, “Nothing about the outrageous costs in publishing is changing. What can I do?” Thankfully Pete Binfield agreed with me and we set out to make public access to research faster, cheaper to produce, and most importantly free to download. No one deserves a punch in the face for pursuing academic knowledge.


CHORUS is now live - how does it stack up to PubMed?

What is CHORUS and why is it important to know about if you’re an academic? From the FAQ (bold emphasis mine):

CHORUS (Clearinghouse for the Open Research of the United States) is a not-for-profit public-private partnership to provide public access to the peer-reviewed publications that report on federally funded research. Conceived by publishers as a public access solution for funding agencies, research institutions, and the public, CHORUS is in active development with more than 100 signatories (and growing). Five goals drive CHORUS’ functionality: identification, discovery, access, preservation, and compliance. CHORUS is an information bridge, supporting agency search portals and enabling users to easily find free, public access journal articles on publisher platforms.

Only it fails in the one thing that it claims to support, public access - at least as far as I can tell so far. And this is the big worry we’ve had all along, that a paywall publisher backed solution to the White House’s OSTP mandate would not work. For a critical overview of the concerns see Michael Eisen’s comments from one year ago when CHORUS was announced.

Why isn’t CHORUS working?

Let us jump right into doing a search. Here’s an example query for NIH funded research. When I ran this search today (August 1, 2014) I got only 3,775 results. Hmmm. That can’t be right, can it? Only 3,775 NIH funded articles? Moving on…

The first result I got was to an article published July 2014 in the American Journal of Medical Genetics. Click the DOI expecting public access, and I hit a paywall. Oh wait, that’s right - CHORUS also indexes embargoed research set to actually be public open access in 12-24+ months. Next several search results - same paywall. Not until the fifth result do I reach an Open Access article.

OK fine. Perhaps it is reasonable to include a mix of embargoed papers with public open access papers - even though OPEN RESEARCH is in the name of CHORUS. I’ll just click the filter for actual public open access papers and see my results. Hmm, unfortunately there is no filter for actual public open access papers. Ruh-rohs. 

And there does not appear to be any labeling on search results indicating whether a paper is actually public open access or still embargoed (for some unknown period of 1-2 years). Ruh-rohs again.

Are we just seeing teething pains here? In some things for sure, for example only having 3,775 NIH results (when there are millions). It can take time to get all of that backlog from publishers (though I don’t know why they’d launch with such a paltry number). However, I don’t believe the lack of Open Access labels or ability to search only for papers already Open Access (rather than embargoed) is a teething problem. That’s a major oversight and makes you wonder why it was left out in a system designed by a consortium of paywall publishers. I can’t imagine SPARC, for example, leaving out an Open Access filter if they had built this search.

What else is wrong with CHORUS? 

The above was just one technical problem, albeit a very concerning one. The main issue is the inherent conflict of interest that exists in allowing subscription publishers the ability to control a major research portal. As Michael Eisen put it, that’s like allowing the NRA to be in charge of background checks and the gun permit database.

In the title I asked, “how does CHROUS stack up to PubMed?” We need to make this comparison since one of the aims of CHORUS is to direct readers to the journal website, instead of reading/downloading from PubMed Central (PMC).

Perhaps most importantly, CHORUS allows publishers to retain reader traffic on their own journal sites, rather than sending the reader to a third party repository.


And if you believe Scholarly Kitchen then PMC is robbing advertising revenues from publishers and PMC is costing taxpayers money as a useless redundant index of actual public/open access papers. Let’s not mince words, Scholarly Kitchen (and by extension the Society for Scholarly Publishing) believes that PubMed and PMC should be shut down. No one believes taxpayer money should be needlessly wasted, but it is a tall order to replace PubMed and PMC, so our expectations for CHORUS should be just as high.

Unfortunately, it is clear from using the CHORUS search tool that I have far less access and insight into publicly available research. And while an open API is slated for the future, it is questionable whether it will be as feature rich as NCBI’s own API into PubMed and PMC. 

CHORUS also fragments an otherwise aggregated index with PubMed. CHORUS looks to index only US-based federally funded research that is either Open Access or slated to be after a lengthy embargo. This means you still need to rely on PMC to find a non-US funded Open Access article. Clearly we still want that since it helps US researchers, right? Then why shut PMC down?

CHORUS isn’t free either. They’ve set the business model up such that publishers pay to have their articles indexed there. Do you think publishers are going to absorb those costs, or pass it along to authors/subscribers? The fact that CHORUS won’t index unless a publisher pays is rather scary; especially if CHORUS were to ever become the defacto database for finding research.

In Summary

I think CHORUS will improve over time, for sure. My worries though are the inherent conflicts of interest and that a major mouthpiece for CHORUS is calling for the removal of PubMed and PMC. I’m also skeptical whenever I see an organization using deceptive acronyms. CHORUS is not a database of Open Research as its name suggests. At least not ‘Open’ in the sense that the US public thinks of open.

You see, if CHORUS can convince the public and US Congress or OSTP that research under a two year embargo is still 'open’ then they’ve won. It’s a setback for what is really Open Access. Nothing short of marketing genius (or manufactured consent) to insert Open Research into the organizational name. 

I think these are legitimate concerns that researchers and the OSTP should be asking of CHORUS.


What have we learned about ourselves from the Eich/Mozilla controversy?

So Brendan Eich has resigned as CEO from Mozilla. From the words of Mozilla Executive Chairwoman Mitchell Baker, this wasn’t a result of his past donation to Proposition 8 in California banning gay marriage, but rather “It’s clear that Brendan cannot lead Mozilla in this setting…The ability to lead — particularly for the CEO — is fundamental to the role and that is not possible here.” i.e. the controversy that this brewed was tarnishing Mozilla’s reputation, trust, brand, etc. 

Unfortunately, this is one of those situations where no one ends up feeling happy. This got ugly on both sides. It is sad that the co-founder of Mozilla and the creator of javascript had to resign. It sucks, I’m sure most of us had high hopes. At the same time Mozilla was being led by someone who wouldn’t apologize for wanting to ban gay marriage, so people had every right to voice disagreement over that promotion. 

Now however, there is a lingering “meta” controversy over whether the way this was handled on both sides was wrong or right. Was the “anti-Eich” crowd too vengeful, too close to being what has been described as being a lynch mob, were they being hypocritical and intolerant? Were the folks supporting Eich being insensitive to a growing civil rights movement, misunderstanding what Mozilla represents, erroneously mixing a specific case with a hypothetical slippery slope?

In the larger picture, these questions and issues exposed during the controversy are just one more signal that “tech” still has much growing up to do. On the one hand tech is dominating every industry and part of our lives. “Software is eating the world,” a quote from Marc Andreessen, which is often used to describe what is going on. And tech is still in its infancy or teenage years in terms of how long it has been with us, which brings challenges as it inundates our culture and organizational behavior practices. Tech is even influencing our politics now, as seen with Twitter in various countries, and the SOPA movement. Some say the tech community went too far with SOPA when websites blacked themselves out in protest, and of course the sexism that continually rears its head in tech is immaturity at its best.

When you have something that is massively influencing every part of our lives, but is still immature, then it can only lead to more “meta” controversies like the Mozilla one. We (i.e. the community, public) simply just don’t know how to appropriately react yet to these situations, it’s going to take time to adjust to what tech in our lives means. Although I think most calling for Eich’s resignation were proportionate in their response, their were outliers who went too far in how they handled it. There will always be people who go too far of course, but now tech can ignite these crowds in a blink of an eye and carry people with it who would not normally participate. That said, I’m confident that as we grow to understand what tech means in our lives that we will resolve that issue in time.

I think we would all be served well if a post-mortem was done for this particular controversy. And this increasingly common situation of how “tech reacts” deserves to be studied as a larger whole. There is an awful lot that we could learn about ourselves and tech in our lives. Important as tech and its social impact isn’t going away, it will only increase.


The most enlightening reveal of the Mozilla CEO controversy

A new exclusive interview with anti-LGBT supporter Brendan Eich on CNET shows that the controversy is not dying down. My own thoughts on why this is a bad choice as CEO for Mozilla Foundation were posted a few days ago. Since then, we’ve seen three board members step down in conflicting reports stating they resigned in a form of protest, contrasting with Eich in the CNET interview and the remaining board stating they were long planned for. 

Needless to say, things are very foggy over at Mozilla and the future is still unclear. One thing is clear, however, that the leadership (CEO and Board) in fact is incapable of leading. Even if they now decide to fire Eich and replace him with a more forward-thinking CEO, it will be only because they’ve caved after sitting back for weeks to measure public opinion; that’s playing politics rather than leading. The Mozilla board I’d like to see is one that knows what the right thing to do is from the get-go or decisively changes course if a mistake is made. What this controversy has revealed is that Mozilla plays politics, it doesn’t lead. That paints a picture where the future at Mozilla will inevitably be filled with more mistakes.

Mozilla doesn’t share my values - both in terms of installing a CEO who doesn’t support LGBT rights and in it’s overall leadership characteristics. It’s disappointing. I’d like to see the remaining board resign and I won’t be returning to any Mozilla products until that happens. 


Mozilla needs to skate to where the puck is going, not where it’s been. LGBT rights

I just don’t understand WTF the Mozilla Foundation was thinking on this one. It’s akin to making someone a CEO who donated to segregation campaigns in the 1960s. You just wouldn’t do that.

Gay rights have not yet achieved the same acceptance as other civil rights have, but they will do so undoubtedly one day. And when that day arrives, and it already has in most of Mozilla’s fan base markets, it is going to haunt the Mozilla Foundation even more than the decision today is doing. Mozilla looks to the future in all it does, except when it comes to its leadership apparently. The future profile of CEOs will not have anti-LGBT beliefs, and Mozilla needs to be skating to where the puck is going, not where it’s been.

What is really troubling is that the origins of the tech industry, the West Coast tech, were about tolerance and civil rights. Personal computers and related technologies were about freedom from oppression. For a tech non-profit to install an anti-LGBT CEO is completely 180 from why and how tech evolved from its early days. Technology isn’t just about what you do, it’s also about who does it. They go hand in hand.

If this was simply an oversight that was missed in due diligence of the CEO’s past then OK, but make the change happen. 


Does the Google Cloud services price drop spell trouble for AWS? It depends.

With Google announcing massive price drops (https://developers.google.com/storage/) it has a lot of developers and tech managers re-thinking the use of AWS, Rackspace, etc. Certainly the $.026/GB of monthly storage and lowered compute engine prices make stiff competition. I am not convinced yet, however. 

First, depending on what one is trying to accomplish, the new bandwidth prices announced by Google are still more expensive than AWS once you reach a certain volume. For heavy bandwidth out users then, it may not make sense to use Google over AWS if pricing is your only concern. 

A larger issue is one of trust, and opinion shows many developers and decision makers are in agreement with me on this one. Google has failed users and developers countless times as it has pulled its APIs and services. Google has a one year notification term for its cloud computing services, but that can change, and one year is possibly not enough time if your entire business is structured around the service. The fine details of the cloud computing terms of use with Google could also give one pause compared to AWS.

Further, AWS revenue accounts for ~7% of Amazon’s overall business revenue. In comparison, the Google cloud computing business brought in roughly 1.5% of Google’s overall 2013 revenue. That alone is enough to give me a second thought when considering trusting my business or computing needs with Google due to its lengthy history of pulling services. 

There are of course other reasons to distrust Google, which I won’t go into now.

AWS is likely to follow Google and continue dropping its prices as well. So, any prudent decision-makers should take a wait-and-see approach before jumping ship. And even then make a careful analysis (including future growth scenarios) of how your business or process actually utilizes the various cloud services to determine the real and future costs involved. 


FIRST Act isn’t the first to use doublespeak against the advancement of science

The Frontiers in Innovation, Research, Science and Technology (FIRST) Act (link) is doublespeak for “we’re actually going to limit Open Access.”

The FIRST Act is yet another bill that is winding its way through the US Congress that despite making claims FOR science will actually reduce the availability of Open Access. Luckily the Scholarly Publishing and Academic Resources Coalition (SPARC) has clarified the damage that this bill would actually do to scientific advancement within the U.S. PLOS has done another writeup of the severe consequences this bill would bring. 

In the past similar bills such as the RWA "Research Works Act" backed by the Association of American Publishers and many paywall publishers have used this doublespeak. The Clearinghouse for the Open Research of the United States (CHORUS), a publisher backed proposal,  is another initiative filled with doublespeak, with the real aim to control access - not open it up. And more recently the “Access to Research” initiative from publishers does the opposite of what its title proclaims. It limits access to research in the digital age by adding a physical barrier and forcing you to travel hundreds of miles to a participating library instead of providing access in the convenience of your lab or home. 

What really fascinates me, however, is the continued use of marketing doublespeak in these legacy publisher proposals to manufacture consent and distort the facts for financial gain. That they are pronounced with a straight face each time makes me just a little sick inside that people like this actually exist. The opposite of heroes, value creators, and leaders. If you haven’t noticed, these tactics grind my gears to the point of evoking a visceral emotional response.

Now I’ve looked to see who outside Congress is backing the FIRST Act by way of either public support or Congressional campaign donations and have yet to find a connection to the usual suspect publishers or associations. Please leave a comment if you do find a connection. 


As Björn Brembs points out, a number of paywall journals and publishers have donated to the Congressmen responsible for bringing the FIRST Act to the House of Reps. This is more than a smoking gun leading back to Elsevier, and a few other large publishers known for backing previous anti-OA bills.